Participants assigned to the intervention group will be given a manual containing instructions to play the Safe City game. A research assistant will provide a briefing session on the game in each participating school. A unique username and password set will be created for each user to log in the game. This login information will be provided to the student participants in a sealed envelope after the briefing session. The participants will be instructed to play the game as many times as desired within a 4-week time frame. The players ranked in the top 20 will receive a reward in the form of book coupon after the intervention ends.

All students in the health education group will receive a comprehensive package on safety information. The information package includes both printed and electronic promotional materials regarding safety and a comprehensive list of relevant website and information sources. The information from these relevant websites and information sources are similar to those used in setting the safety case scenarios for the Safe City game. As a result, both intervention arms will have comparable accessibility to safety-related information and the major contrast between the two groups will be the method of presentation (game-based learning vs traditional health promotion approach, i.e. unidirectional information package).

The game intervention uses a game-based learning approach to teach players how to recognize and remove injury hazards from different homes and sports areas. It also highlights the need to pay attention to traffic and safety facilities to prevent injuries in a real-world environment. Players will create their own avatar to assume the duties of safety inspectors in the game. Point, level and reward systems will be included to increase player motivation to engage in the game.

Child safety knowledge will be measured by different items adopted from the existing literature and questionnaires under the three contexts: home safety, road safety, and sports-related safety. Child participants will be asked to indicate their level of involvement in the said behaviour on a 5-point Likert scale ranging from "Never" to "Very Often", or their agreement on the given statement based upon their own safety knowledge on a 4-point Likert scale ranging from "Strongly disagree" to "Strongly agree".

Child safety knowledge will be measured by different items adopted from the existing literature and questionnaires under the three contexts: home safety, road safety, and sports-related safety. Child participants will be asked to indicate their level of involvement in the said behaviour on a 5-point Likert scale ranging from "Never" to "Very Often", or their agreement on the given statement based upon their own safety knowledge on a 4-point Likert scale ranging from "Strongly disagree" to "Strongly agree".

Child safety behaviours will be measured by different items adopted from the existing literature and questionnaires under the three contexts: home safety, road safety, and sports-related safety. Child participants will be asked to indicate their level of involvement in the said behaviour on a 5-point Likert scale ranging from "Never" to "Very Often", or their agreement on the given statement based upon their own safety knowledge on a 4-point Likert scale ranging from "Strongly disagree" to "Strongly agree".

Child safety behaviours will be measured by different items adopted from the existing literature and questionnaires under the three contexts: home safety, road safety, and sports-related safety. Child participants will be asked to indicate their level of involvement in the said behaviour on a 5-point Likert scale ranging from "Never" to "Very Often", or their agreement on the given statement based upon their own safety knowledge on a 4-point Likert scale ranging from "Strongly disagree" to "Strongly agree".

Child generic quality of life will be measured by the Chinese version of the PedsQL Generic Core Scales (Child-Self Report version) which has four subscales with a total of 23 items measuring the physical, emotional, social, and school functioning of the children. Each item is rated on a 5-point scale from 0=never to 4=almost always. A higher score indicates better functioning.